Overview

This is a simple package to make my work easier, with some encapsulated functions to achieve daily common analysis with a uniform interface.

analyze function

All common analyses can be implemented by function analyze, by specifying the object and assigning task to the object. A concrete example is as follows:

library(org.Hs.eg.db)
data(gene_vector)
y <- analyze(object = gene_vector, task = 'go')

Here it will automatically call the corresponding function to achieve analysis using some default arguments. The analysis’s details can be adjusted by extra arguments if you customize some of the arguments. For example, you can customize the ont argument to analyze all GO categories analysis:

y <- analyze(object = gene_vector, task = 'go', ont = 'ALL')

If you want to do KEGG analysis, just assign task to kegg.

data(kegg)
y <- analyze(object = kegg, task = 'kegg')

plot function

All analyses’ results can be visualized using plot function which extends the plot generic in the base package. you can easily visualize the results just like this:

p <- plot(y)

The plot function returns a ggplot2 object which is easy to adjust the details of the figures according to the user’s requirements.

library(viridis)
p <- p + scale_fill_viridis(discrete = true)

Object Reductor

Reductor is an R6 object which is mainly used to view the results of dimensional reduction in different arguments combinations.

library(palmerpenguins)

x <-
  penguins |>
  tidyr::drop_na()

y <- x$species
x <- x |>
  dplyr::select(where(is.numeric)) |>
  dplyr::select(-year)
x <- scale(x)
# show the results of different arguments
happy <- Reductor$new("tsne")
set.seed(20230530)
tune_fit <- happy$tune(x,
  perplexity = c(30, 40, 50, 60),
  n_iter = c(1000, 2000, 2500)
)
happy$plot(tune_fit, y)